The future of engineering

Google and NYU Tandon join forces to discuss where we are now and where we’re headed

Panelists on stage

Phil Venables, Chief Information Security Officer at Google; Juliana Freire, Professor of Computer Science & Data Science; Miguel Modestino, Associate Professor of Chemical Engineering; Todd Underwood, Senior Engineering Director at Google; Nasir Memon,  Professor of Computer Science & Engineering and Jelena Kovačević, NYU Tandon Dean 

What’s the use of studying linear algebra? What does engineering have to do with philosophy? Is knowing all about language models enough to forge a career in AI?

On March 23, Tandoneers filled Pfizer Auditorium to hear a panel discussion featuring Google executives and NYU Tandon faculty members touch upon those topics and many more as they examined the current and future state of engineering education and the job market. (The answers to the above: Linear algebra represents foundational knowledge; philosophers offer critical insight into responsible engineering, and Google has hired them to consult on ethical issues; and no, decidedly not.) 

Moderated by Tandon Dean Jelena Kovačević, the panel included:

While the panelists were jovial and the discussion free-roaming, the major focus was on serious issues of great relevance to students, their professors, and the employers who may one day hire them — namely, with tech evolving so fast, which problems should the tech world be tackling and how should engineering education contribute? 

There was broad consensus that theory and practice are both important — a fact that engineering schools must take into consideration as curriculum is designed and the balance between teaching principles and providing experiential learning opportunities is weighed–and that it isn’t enough to understand the technology you are building: you must understand its role in society and impact on users. It’s relatively easy, Venables asserted, to create a product; the real challenge is how to introduce that product safely and securely.

Safety and security were top of mind for all the panelists, including Memon, who pioneered one of the first cybersecurity degree programs in the nation. He pointed out that every tech job now requires an understanding of security aspects, whether you are a coder developing what you hope is an impenetrable system or an administrator who must know at least enough to keep your company credentials protected — even though decades ago, STEM professionals felt little need to consider security and privacy as priorities. 

It’s thus no longer enough to train cybersecurity professionals; engineering schools need to ingrain security principles into the fabric of engineering education across disciplines. That holds for other principles as well: it’s not only environmental engineers, for example, who must keep sustainability in mind. (It would not have been standard procedure to consider the power consumption of computing before this, but in an era where AI technology is proving to have an enormous carbon footprint, that now needs to become a key factor in weighing new advances in algorithmic models and rolling out new AI systems.)

Because other cross-cutting concerns will undoubtedly emerge in the years to come, it’s also not enough, panelists explained, to simply complete a course of study and consider yourself educated: Everyone must maintain solid foundational knowledge (that’s where the linear algebra comes in), remain a lifelong learner, and stay flexible, since, as Underwood put it, “the future is less predictable than you think.” 

Watch highlights from the panel discussion

Modestino and Friere stressed that learning to harness the massive amounts of data now being generated was also a key to success in any number of fields. “I’m computationally competent but I’m not a computer scientist, I’m a chemical engineer,” Modestino said. “And now we are moving away from just chemical intuition to use the data we collect to inform that intuition. Ultimately the foundational knowledge remains critical, but with the technology becoming more accessible, it presents more opportunities to do more things, more quickly.”

Friere elaborated: “We’ve witnessed a perfect storm: computing and storage have become virtually free, and we have lots and lots of data,” she said. “This has changed how science is done — people have moved from hypothesis-driven science to using data and computing to generate new hypotheses.” Engineering schools must therefore try to ensure that those in a broad range of fields are able to glean actionable insights from data — admittedly a difficult task even for computer and data scientists. “The big challenge I see is how we democratize data and computing to a broader audience,” Friere asserted. “There’s so much more that’s needed than language models. We need data engineering skills and data literacy.”

With the job market of deep concern to students, the panelists spent some time reassuring them that the hiring landscape was less bleak than many headlines implied. Fields adjacent to computer science are booming, they explained, and there are job prospects everywhere if you are willing to explore all the verticals. (Google itself places emphasis on the interdisciplinary, meaning that the company needs well-rounded engineers comfortable working with lawyers and philosophers.)

Responding to fears that ChatGPT is set to replace developers altogether, Underwood explained that the platform will probably at some point take over the more tedious coding and testing tasks–the ones few people actually want to do anyway–but that humans would always be needed for the more challenging, creative tasks.

Modestino, who mentored 2019 Global Student Entrepreneur of the Year Danielle Blanco, among other aspiring business founders, reminded the audience that no matter what the future holds, they don't have to look for a job: they can create one for themselves. It was one final bit of good advice in an event filled with useful, hopeful (and fun) information.